An ion source is a device that ionizes gas molecules and then, accelerates, focuses, and emits them as an ion beam. This beam is then used for various technical and technological purposes such as cleaning, activation, polishing, thin-film coating, or etching.
An example of an ion source is the so-called Kaufman ion source, also known as a Kaufman ion engine or an electron-bombardment ion source described in U.S. Pat. No. 4,684,848 issued to H. R. Kaufman in 1987.
This ion source consists of a discharge chamber, in which a plasma is formed, and an ion-optical system which generates and accelerates an ion beam to an appropriate level of energy. A working medium is supplied to the discharge chamber which contains a hot cathode that functions as a source of electrons and is used for firing and maintaining a gas discharge. The plasma, which is formed in the discharge chamber, acts as an emitter of ions and creates, in the vicinity of the ion-optical system, an ion-emitting surface. As a result, the ion-optical system extracts ions from the aforementioned ion-emitting surface, accelerates them to a required energy level, and forms an ion beam of a required configuration. Typically, aforementioned ion sources utilize two-grid or three-grid ion-optical systems.
A disadvantage of such a device is that it requires the use of ion accelerating grids and an ion beam of low intensity. Another disadvantage is that this apparatus is not suitable for operation with reactive gases.
Attempts have been made to provide ion sources with ion beams of higher intensity by holding the electrons in a closed space between a cathode and an anode where the electrons could be held. For example, U.S. Pat. No. 4,122,347 issued in 1978 to Kovalsky et al. describes an ion source with a closed-loop trajectory of electrons for ion-beam etching and deposition of thin films, wherein the ions are taken from the boundaries of a plasma formed in a gas-discharge chamber with a hot cathode. The ion beam is intensified by a flow of electrons which are held in crossed electrical and magnetic fields within the accelerating space and which compensate for the positive spatial charge of the ion beam.
A disadvantage of devices of such type is that they do not allow formation of ion beams of chemically-active substances because of the use of a hot cathode and because of strong localization of the plasma-formation zone. Other disadvantages of the aforementioned devices are short service life and high non-uniformity of ion beams.
U.S. Pat. No. 4,710,283 issued in 1997 to Singh et al. describes a coldcathode type ion source with crossed electric and magnetic fields for ionization of a working substance wherein entrapment of electrons and generation of the ion beam are performed with the use of a grid-like electrode. This source is advantageous in that it forms belt-like and tubular ion beams emitted in one or two opposite directions.
However, the ion source with a grid-like electrode of the type disclosed in U.S. Pat. No. 4,710,283 has a number of disadvantages consisting in that the grid-like electrode makes it difficult to produce an extended ion beam and in that the ion beam is additionally contaminated as a result of sputtering of the material from the surface of the grid-like electrode. Furthermore, with the lapse of time the grid-like electrode is deformed whereby the service life of the ion source as a whole is shortened.
Other publications (e.g., Kaufman H. R. et al. (End Hall Ion Source, J. Vac. Sci. Technol., Vol. 5, Jul/Aug., 1987, pp. 2081-2084; Wykoff C.A. et al., 50-cm Linear Gridless Source, Eighth International Vacuum Web Coating Conference, Nov. 6-8, 1994)) disclose an ion source that forms conical or belt-like ion beams in crossed electrical and magnetic fields. The device consists of a cathode, a hollow anode with a conical opening, a system for the supply of a working gas, a magnetic system, a source of electric supply, and a source of electrons with a hot cathode. A disadvantage of this device is that it requires the use of a source of electrons with a hot or hollow cathode and that it has electrons of low energy level in the zone of ionization of the working substance. These features create limitations for using chemically-active working substances. Furthermore, a ratio of the a cathodeanode distance to the Larmor radius is significantly greater than 1, and this decreases the energy of electrons in the charge gap, and hence, hinders ionization of the working substance. Configuration of the electrodes used in the ion beam of such sources leads to a significant divergence of the ion beam and degradation of the cathode material because of sputtering. As a result, the ion beam cannot be delivered to a distant object and is to a greater degree subject to contamination with the material of the electrode. In other words, the device described in the aforementioned literature is extremely limited in its capacity to create an extended uniform belt-like ion beam. Another disadvantage for target sputtering is a low energy of ions which does not allow effective sputtering of nonconductive, e.g., dielectric materials.
Russian Patent No. 2,030,807 issued in 1995 to M. Parfenyonok, et al. describes an ion source that comprises a magnetoconductive housing used as a cathode having an ion-emitting slit, an anode arranged in the housing symmetrically with respect to the emitting slit, a magnetomotance source, a working gas supply system, and a source of electric power supply.
FIGS. 1 and 2 schematically illustrate the aforementioned known ion source with a circular ion-beam emitting slit. More specifically, FIG. 1 is a sectional side view of an ion-beam source with a circular ion-beam emitting slit, and FIG. 2 is a sectional plan view along line II--II of FIG. 1.
The ion source of FIGS. 1 and 2 has a hollow cylindrical housing 40 made of a magnetoconductive material such as Armco steel (a type of a mild steel), which is used as a cathode. Cathode 40 has a cylindrical side wall 42, a closed flat bottom 44 and a flat top side 46 with a circular ion emitting slit 52.
A working gas supply hole 53 is formed in flat bottom 44. Flat top side 46 functions as an accelerating electrode. Placed inside the interior of hollow cylindrical housing 40 between bottom 44 and top side 46 is a magnetic system in the form of a cylindrical permanent magnet 66 with poles N and S of opposite polarity. An N-pole faces flat top side 46 and S-pole faces bottom side 44 of the ion source. The purpose of a magnetic system 66 with a closed magnetic circuit formed by parts 66, 40, 42, and 44 is to induce a magnetic field in ion emitting slit 52. It is understood that this magnetic system is shown only as an example and that it can be formed in a manner described, e.g., in aforementioned U.S. Pat. No. 4,122,347. A circular annular-shaped anode 54 which is connected to a positive pole 56a of an electric power source 56 is arranged in the interior of housing 40 around magnet 66 and concentric thereto. Anode 54 is fixed inside housing 40 by means of a ring 48 made of a non-magnetic dielectric material such as ceramic. Anode 54 has a central opening 55 in which aforementioned permanent magnet 66 is installed with a gap between the outer surface of the magnet and the inner wall of opening 55. A negative pole 56b of electric power source is connected to housing 40 which is grounded at GR.
Located above housing 40 of the ion source of FIGS. 1 and 2 is a sealed vacuum chamber 57 which has an evacuation port 59 connected to a source of vacuum (not shown). An object OB to be treated is supported within chamber 57 above ion emitting slit 52, e.g., by gluing it to an insulator block 61 rigidly attached to the housing of vacuum chamber 57 by a bolt 63 but so that object OB remains electrically and magnetically isolated from the housing of vacuum chamber 57. However, object OB is electrically connected via a line 56c to negative pole 56b of power source 56. Since the interior of housing 40 communicates with the interior of vacuum chamber 57, all lines that electrically connect power source 56 with anode 54 and object OB should pass into the interior of housing 40 and vacuum chamber 57 via conventional commercially-produced electrical feedthrough devices which allow electrical connections with parts and mechanisms of sealed chambers without violation of their sealing conditions. In FIG. 1, these feedthrough devices are shown schematically and designated by reference numerals 40a and 57a. Reference numeral 57b designates a seal for sealing connection of vacuum chamber 57 to housing 40.
The known ion source of the type shown in FIGS. 1 and 2 is intended for the formation of a unilaterally directed tubular ion beam. The source of FIGS. 1 and 2 forms a tubular ion beam IB emitted in the direction of arrow A and operates as follows.
Vacuum chamber 57 is evacuated, and a working gas is fed into the interior of housing 40 of the ion source. A magnetic field is generated by magnet 66 in the accelerating gap between anode 54 and cathode 40, whereby electrons begin to drift in a closed path within the crossed electrical and magnetic fields. A plasma 58 is formed between anode 54 and cathode 40. When the working gas is passed through the ionization gap, tubular ion beam EB, which is propagated in the axial direction of the ion source shown by an arrow A, is formed in the area of an ion-emitting slit 52 and in an accelerating gap 52a between anode 54 and cathode 40.
The above description of the electron drift is simplified to ease understanding of the principle of the invention. In reality, the phenomenon of generation of ions in the ion source with a closed-loop drift of electrons in crossed electric and magnetic fields is of a more complicated nature and consists in the following.
When, at starting the ion source, a voltage between anode 54 and cathode 40 reaches a predetermined level, a gas discharge occurs in anode-cathode gap 52a. As a result, the electrons, which have been generated as a result of ionization, begin to migrate towards anode 54 under the effect of collisions and oscillations. After being accelerated by the electric field, the ions pass through ion-emitting slit 52 and are emitted from the ion source. Inside the ion-emitting slit, the crossed electric and magnetic fields force the electrons to move along closed cycloid trajectories. This phenomenon is known as "magnetization" of electrons. The magnetized electrons remain drifting in a closed space between two parts of the cathode, i.e., between those facing parts of cathode 40 which form ion-emitting slit 52. The radius of the cycloids is, in fact, the so-called doubled Larmor radius & which is represented by the following formula: EQU R.sub.L =mcV/.vertline.e.vertline.B,
where m is a mass of the electron, B is the strength of the magnetic field inside the slit, V is a velocity of the electrons in the direction perpendicular to the direction of the magnetic field, and .vertline.e.vertline. is the charge of the electron.
It is required that the height of the electron drifting space in the ion-emission direction be much greater than the aforementioned Larmor radius. This means that, in distinction from the aforementioned Hall-end Kaufman ion source, where a ratio of cathode-anode distance to the Larmor radius is equal approximately to 1. In other words, a part of the ionization area penetrates into ion-emitting slit 52 where electrons can be maintained in a drifting state over a long period of time. In other words, a spatial charge of high density is formed in ion-emitting slit 52.
When a working medium, such as argon which has neutral molecules, is injected into the slit, the molecules are ionized by the electrons present in this slit and are accelerated by the electric field. As a result, the thus formed ions are emitted from the slit towards the object. Since the spatial charge has high density, an ion beam of high density is formed. This beam can be converged or diverged by known technique for specific applications.
Thus, the electrons do not drift in a plane, but rather along cycloid trajectories across ion-emitting slit 52. However, for the sake of convenience of description, here and hereinafter such expression as "electron drifting plane" or "drifting in the direction of ion-beam propagation" will be used.
The diameter of the tubular ion beam formed by means of such an ion source may reach 500 mm and more.
The ion source of the type shown in FIG. 1 is not limited to a cylindrical configuration and may have an elliptical or an oval-shaped cross section as shown in FIG. 3. FIG. 3 is a cross-sectional view of the ion-beam source along line III--III of FIG. 1. In FIG. 3 the parts of the ion beam source that correspond to similar parts of the previous embodiment are designated by the same reference numerals with an addition of subscript OV. Structurally, this ion source is the same as the one shown in FIG. 1 with the exception that a cathode 40.sub.ov, anode 54.sub.ov, a magnet 66.sub.ov, and hence an emitting slit (not shown in FIG. 3), have an oval-shaped configuration. As a result, a belt-like ion beam having a width of up to 1400 mm can be formed. Such an ion beam source is suitable for treating large-surface objects when these objects are passed over ion beam IB emitted through emitting slit 52.
With 1 to 3 kV voltage on the anode and various working gases, this source makes it possible to obtain ion beams with currents of 0.5 to 1A. In this case, an average ion energy is within 400 to 1500 eV, and nonuniformity of treatment over the entire width of a 1400 mm-wide object does not exceed .+-.2%.
Nevertheless, the aforementioned belt-type ion source is disadvantageous in that for use in sputtering, this ion source also requires the use of spatially located targets of sputterable material. This increases the overall dimensions of a sputtering system. A sputtering system of such type is disclosed in pending U.S. patent application Ser. No. 09/161,581 of the same applicants filed in Sep. 28, 1998 now U.S. Pat. No. 6,130,507 issued Oct. 10, 2000.
The above problems are partially solved by a device which in the sputtering technology is known as a magnetron plasma source. Such a device is described, e.g., by J. Reece Roth in "Industrial Plasma Engineering", Institute of Physics Publishing, Bristol and Philadelphia, 1995, p. 337. Basically, a magnetron plasma source, which hereinafter will be referred to as "sputtering magnetron", incorporates a crosswise magnetic field over the cathode, which traps the beam electrons in orbits in that location and thus greatly increases their path length before they finally escape to the anode by collisional scattering. Because the electron's travel path becomes longer than the electron gap, the minimum pressure to sustain the plasma is much lower for the sputtering magnetron than, e.g., for a planar diode--typically 0.1 Pa instead of 3 Pa. At 0.1 Pa, the sputtered particles retain most of their kinetic energy upon reaching the substrate, so the one obtains the beneficial effects of this energy. A planar diode is a plasma source device which consists of two parallel electrodes, namely cathode and anode, where the cathode is grounded and the anode is under a high positive potential with respect to the ground. If a plasma is generated in a cathode-anode space and an object is placed into this plasma, the material of the cathode will be deposited on the object.
A typical sputtering magnetron construction is shown in FIG. 4 which is a sectional three-dimensional view of a planar-magnetron structure. In this drawing, for clarity, the electron-orbit radius is shown much larger than the actual size.
The device consists of a vacuum chamber 100 the walls of which function as an anode. The bottom of the chamber is formed by a watercooled copper backing plate 102 which supports a target 104 of a sputterable material in the form of a 3 to 10 mm thick disk which is bonded for good thermal contact to backing plate 102. Bonding can be done by soldering, or through an epoxy resin, but preferably through clamping in order to provide possibility for replacing sputterable target 104 after its consumption. Vacuum chamber 100 is sealed by an insulating ring 105, e.g., of ceramic, placed between the lower end of vacuum chamber 100 and the flange of the copper backing plate 102. The vacuum chamber wall or an anode assembly is grounded at G. The cathode assembly is formed by copper backing plate 102, a ring of bar magnets 106 and one central magnet 108 inserted into the respective recesses 110 and 112 of backing plate 102, and a Fe field return plate 114 for completion of the magnetic field. The cathode is under a high negative potential, e.g., of about -400 to -800 V. Using the strongest magnets (Nd--Fe--B), the field over the target can approach 1 KGa, or 0.1 T in SI units.
Upon igniting the plasma, flow of electrons emitted from the cathode become accelerated into the plasma by the cathode-sheath electric field E, just as in the case of a planar diode. The presence of the magnetic field B, however, causes them also curve into orbits, and the plasma electrons are magnetized, as was described earlier. The ions, however, are not magnetized. The sputtering magnetron will still operate as a sputtering source at much higher pressure, but gas scattering will dominate electron behavior rather than B.
At low pressure, the sputtering magnetron operates as follows: electrons emitted from the sputterable target surface or created by ionization in the sheath field are accelerated vertically by E but, at the same time, forced sidewise by B, so they eventually reverse direction and return toward sputterable target 104, decelerating in E as they proceed until their direction is again reversed and the cycle repeat. As has by described earlier, the electrons would follow a "cycloidal" path. In reality, the path is more complicated because of collision and because E decreases with distance from sputterable target 104.
Also, deposition rate is increased in the sputtering magnetron because of reduced scattering and redeposition of sputtered particles on the cathode. Finally, the increased efficiency of electron usage means that lower applied voltage (typically 500 V) is needed to sustain a plasma of a given density, and that the voltage increases even less steeply with power than it does in the planar diode.
Magnetrons can be energized by direct-current or high-frequency (RF) power sources. A main disadvantage of a sputtering magnetrons of both type is high pressure of the working medium, high ignition voltage of discharge, high discharge-sustaining power a significant part of which is converted into the Joule heat on the target and which require intensive cooling. Although the RF magnetrons are capable of sputtering nonconductive materials such as dielectric, direct-current magnetrons cannot sputter nonconductive materials. Furthermore, in a magnetron the pressure of the working medium is permanent inside the entire vacuum chamber, i.e., near the sputterable target and near the object (substrate) being treated. The magnitude of this pressure is selected so as to ensure the maximum efficiency of the working gas ionization. This pressure, however, is not always optimum for treating the substrate surface. In other words, the optimum pressure required for surface treatment and the pressure required for optimization of the working gas ionization are contradictory. Another disadvantage of the sputtering magnetron is that the erosion pattern of sputtering magnetron target is highly nonuniform across the target surface. A serious disadvantage of a magnetron is that a dielectric film is often formed on the sputtering surface of target in the case of a reactive sputtering. This film decreases the sputtering surface and is subject to electric breakdown with erruption of particles of target having dimensions much greater than the sputtered ions. This contaminates the object being treated.
It is known to combine an ion beam source with a sputtering magnetron in a vacuum chamber for enhancing sputtering. Such a device is described in U.S. Pat. No. 5,618,389 issued in 1996 to K. Kreider. The device is used for producing transparent carbon nitride films. The films are made by using a magnetron sputter gun and an ion beam source in a vacuum chamber. A disadvantage of this device is that it employs two separate and complicated units, i.e., an ion beam source and a sputtering magnetron. This makes the device expensive and with large overall dimensions. The application is limited because the direction of sputtering by the magnetron and by the ion source are different, i.e., the emission occurs under different angles to the surface of the object. However, the geometry of an object being treated not always allows for utilization of two separate processing units, i.e., the ion beam source and the sputtering magnetron emits ion beams or sputter particles under different angles with respect to the surface of the object. Each of these units requires the use of separate power supply sources, gas delivery systems, etc. Optimum modes of operation (gas pressure, gas flows, etc.) are not always compatible for the use of the sputtering magnetron and the ion source simultaneously in a common vacuum chamber. Furthermore, the apparatus of the aforementioned patent does not provide formation of coating films of uniform thickness and properties.
The problems of the apparatus of U.S. Pat. No. 5,618,389 are partially solved by an apparatus disclosed in U.S. Pat. No. 6,153,067 issued Nov. 28, 2000. now U.S. Pat. No. 6,153,067. This apparatus combines a cold-cathode ion source with a closed-loop ion-emitting slit and a target-sputtering direct-current magnetron in which the direction of propagation of the ion beam coincides with the direction of the flow of sputtered particles. In this apparatus, the sputterable target is placed either onto the entire cathode or onto a part thereof which is charged negatively with respect to another part which is grounded. During the operation of the apparatus, the ion beam emitted from the ion source acts as a virtual anode with respect to the magnetron target. More specifically, in operation of this apparatus, a negative potential of about -500V is applied from a power supply source to a cathode assembly and thus to a sputterable target, while a positive voltage of about 1000V to 1500V is applied to the anode, and a magnetic field B of about 1000 to 3000 Ga is generated across the ion-emitting slit. The potential difference of 500 to 1500 V between the anode and the cathode assembly causes a discharge in the anodecathode space. When working gas passes via the anode-cathode space, a discharge occurs with the formation of the plasma. The molecules of the gas are ionized. The ions are accelerated by the electric field E, and an ion beam IB is formed, which has a charge positive with respect to the sputterable target. The ion beam IB itself has a positive potential with respect to the object being treated, i.e., with respect to the ground, since the object is grounded. This positive potential is about several hundred volts. On the other hand, the cathode assembly and the sputterable target are under a negative voltage of about -500V. Thus the cathode/taget and the ion beam function as a sputtering magnetron where the positive ion beam functions as a virtual anode, transparent to electrons and thus being for them a "potential well". Ions formed in the magnetron discharge are accelerated toward the sputterable target, bombard the target, thus sputter its material, and knock-out secondary electrons from the target. These secondary electrons are accelerated by the negative potential of the target, are held in the crossed electric and magnetic fields of the target, and ionize molecules of the working medium. During operation of the magnetron, this process occurs constantly. The sputtered particles of the target are deposited onto the object forming a coating film. The magnetron-sputtering mode described above occurs simultaneously with the operation of the same apparatus as a pure ion source which enhances the treatment. This is because the combined ion-beam emission and magnetron sputtering significantly increases the life of high-velocity electrons and thus improves efficiency of ionization of the working medium. As a result, it becomes possible to reduce the working pressure in the magnetron sputtering discharge, decrease the discharge voltage, and improve quality of the film being deposited. The quality is improved because the residual gas is to a lower degree contaminates the coating film.
Nevertheless, the apparatus of aforementioned U.S. Pat. No. 6,153,067 has a number of disadvantages. A main disadvantage is that the sputterable target should have a special geometry which is determined by the shape of the cathode and may consist of two parts to cover the inner and outer parts of the upper plate of the cathode. Another problem is associated with cooling, since at high discharge power, the target and cathode require simultaneous intensive cooling.